from json_parser import JsonParser result = JsonParser(data, ["first_name", "street", "address*"]).get_data() CODE_BLOCK: from json_parser import JsonParser result = JsonParser(data, ["first_name", "street", "address*"]).get_data() CODE_BLOCK: from json_parser import JsonParser result = JsonParser(data, ["first_name", "street", "address*"]).get_data() COMMAND_BLOCK: from jsonpath_ng import parse first_names = parse('$..first_name').find(data) streets = parse('$..street').find(data) addresses = parse('$..address*').find(data) # Note: * works but not exactly the same COMMAND_BLOCK: from jsonpath_ng import parse first_names = parse('$..first_name').find(data) streets = parse('$..street').find(data) addresses = parse('$..address*').find(data) # Note: * works but not exactly the same COMMAND_BLOCK: from jsonpath_ng import parse first_names = parse('$..first_name').find(data) streets = parse('$..street').find(data) addresses = parse('$..address*').find(data) # Note: * works but not exactly the same COMMAND_BLOCK: pip install json-key-parser COMMAND_BLOCK: pip install json-key-parser COMMAND_BLOCK: pip install json-key-parser - JSON structure is unpredictable or changes often (third-party APIs, logs, scraped data) - You just want a flat dict/list of the fields you care about - Duplicate keys appear at different nesting levels (it merges them intelligently) - You hate writing defensive data.get("a", {}).get("b", []) chains - You need filters: $.books[?(@.price > 20)] - You want to update or delete values in place: expr.update(data, new_value) - You need arithmetic, regex, slicing, or parent references - You’re doing metaprogramming (it exposes a clean AST) - You want full paths back: match.full_path - json-key-parser: pure stdlib, < 50 KB installed - jsonpath-ng: robust parser (no longer pulls in ply at runtime in recent versions), still tiny - jsonpath-ng is mature, heavily tested, and used in production everywhere. - json-key-parser is brand new (Beta status) but laser-focused, so its recursion is extremely fast for the 95% use case. - You’re writing ETL scripts, API clients, or data-cleaning notebooks - Your JSON is messy and you just want “first_name, email, total” regardless of nesting - You value readability and minimal code - You need conditional filtering or data transformation - You’re building a library or tool that requires precise, repeatable queries - You want to modify the JSON structure in place